Formhavn
Daylight

Designing for the darkest week of the year

Abstract illustration of a colonnade with a large circular opening

Every December we go back to the buildings we finished five years ago and take light readings at eleven in the morning. It is the least glamorous ritual in the studio and the one that has changed our drawings the most. Almost every problem we find was visible in the first plan, and almost none of it was visible in the photographs we published.

The reason is simple. A building is photographed in June, on a bright day, with the lights off. It is lived in during February, at four in the afternoon, with the lights on and the heating running. Those are two different buildings.

The June problem

Our clients rarely ask for a specific daylight factor. They ask for a home that feels generous, or a library where people stay longer than they planned. Both requests are really about light, and both are decided before the first interior render exists.

The June problem is that a deep plan with a large south window will look wonderful in a summer photograph. In winter, the same room will be bright for two hours and gloomy for six. The window was never the issue. The depth of the room behind it was.

A room is not lit by the window it has. It is lit by the distance between that window and the furthest thing you want to see.

Ingrid Halvorsen, founding partner

Once we started measuring that distance, most of our arguments about windows disappeared. A room seven metres deep needs light from two sides or a roof, whatever the elevation looks like.

Four measurements we always take

Before a plan is approved internally, the project architect brings four numbers to the Monday review. They are not a simulation result; the simulation comes later. They are a quick sanity check that costs twenty minutes.

  1. Room depth against ceiling height. Anything deeper than twice the ceiling height needs a second light source.
  2. Glazing ratio by room. Window area as a percentage of floor area, checked room by room rather than for the building as a whole.
  3. Orientation spread. How many rooms receive light from more than one direction.
  4. Obstruction angle. What stands opposite the window, and at what angle it blocks the winter sun.

The third number is the one that surprises clients. A flat where every room faces the same way is a flat with one mood. Two orientations give you a morning room and an evening room, and that is worth more than an extra four square metres.

The plan test

We use a simple table in every project, filled in at stage two and revisited at stage four. It is deliberately coarse. Precision at this stage is a way of avoiding decisions.

Daylight targets used at concept stage, measured on the darkest week.
Space Daylight factor Our rule
Kitchen 2.0% Glazing at least 12% of floor area
Living room 2.5% Two orientations where the site allows
Bedroom 1.5% No light borrowed from circulation
Circulation 1.0% Borrowed light acceptable
The targets are stricter than the national building regulation, which sets a single figure for the whole dwelling rather than for the rooms people actually use.

Room by room, not building by building

A dwelling can pass a whole-building average while containing one room that is effectively a cupboard with a bed in it. We stopped reporting averages in 2019. Every room is listed, every room is checked, and the one that fails gets redesigned before the plan goes to the client.

What it costs to get it wrong

Lighting a gloomy room artificially is cheap to install and expensive to live with. In a 90 square metre apartment, correcting a dark living room usually means one of three things: a larger opening, a different plan, or a light well. The first is the cheapest and the last is the most disruptive, and all three are decided in week three or not at all.

We keep a small script that converts a room's dimensions and glazing into a first-pass estimate, mostly so that the conversation starts with numbers rather than adjectives:

{
  "room": "living",
  "depth_m": 6.4,
  "ceiling_m": 2.7,
  "glazing_ratio": 0.14,
  "orientations": ["south", "west"],
  "target_df": 2.5,
  "result": "pass"
}

The result field is intentionally blunt: pass or fail. Anything in between tends to be argued about for a week and then built anyway.

A short checklist

Rule of thumb If you need the light on at eleven in the morning in December, the window is in the wrong wall — or the room is too deep for the window it has.

Before signing off any plan, ask four questions:

  • Can I describe every room's light in one sentence without mentioning a render?
  • Does any room depend on a wall that will be built by someone else?
  • Have I tested the plan on the shortest day, not the longest?
  • If the answer is no to any of these, who is going to fix it, and when?

None of this is novel. It is the ordinary discipline of drawing a section before falling in love with a plan, and it is the cheapest design work we do. The buildings that feel calm in February are rarely the ones with the largest windows. They are the ones where somebody measured the distance from the window to the far wall and decided to move it.

Leave a comment

Comments are moderated by the studio. We publish criticism as readily as praise.

Never published.
At least 20 characters.

Palette